A systematic review of data-driven learning research on language learning and teaching for pre-tertiary learners: Balancing qualitative and quantitative research

Research has insufficiently addressed young learners' perceptions and the practical implementation of data-driven learning (DDL) at the pre-tertiary level. To address these gaps, we synthesized findings from 45 studies published between 1994 and 2025 using a mixed-methods approach, focusing on two perspectives: the suitability of DDL for young learners and learner anxiety. First, we provided a descriptive map of DDL characteristics at the pre-tertiary level. Second, through quantitative synthesis of 34 studies, we confirmed the effectiveness and efficiency of DDL in second language acquisition across various age groups and types of corpus interaction. Third, through a qualitative synthesis of 26 studies, we identified mostly positive perspectives among young learners, being consistent across all age groups and corpus interaction types. We confirm the suitability of DDL for young learners in pre-tertiary education, provide recommendations for educators, and propose directions for future research on young learners' language education through DDL. In this systematic review, we comprehensively analyzed the application of data-driven learning in non-university settings by synthesizing data from 45 studies published between 1994 and 2025. Our findings confirm that data-driven learning is effective and efficient for young learners across different age groups and corpus interaction types, and most young learners have a positive attitude toward data-driven learning. We also expand the scope of data-driven learning by demonstrating that age does not significantly affect learning outcomes, thereby showing that it can be applied across different educational levels and age groups. Additionally, our review highlights the role of the corpus interaction type in affecting learners' engagement and outcomes. We offer practical recommendations to educators for the design of personalized, corpus-based lessons that satisfy individual learner needs, learning preferences, and experiences. Finally, we advocate for the adoption of data-driven learning at the pre-tertiary level to enhance learners' language proficiency and cognitive abilities in an engaging and effective manner. • Systematic review of DDL for pre-tertiary learners; • Mixed-methods approach synthesizing 45 studies from 1994 to 2025; • Quantitative synthesis confirms effectiveness and efficiency of DDL across different age groups various types of corpus interaction; • Qualitative synthesis identifies positive perceptions of DDL across all age groups and corpus interactions types;